Research On Transmission Robustness And Optimization In Heterogeneous Networks Based On Multipath TCP

With the continuous breakthroughs and innovations in the new generation of Internet technologies,streaming services such as mobile UHD video have become an indispensable part of people’s daily life,while also posing new challenges to the transmission performance of mobile communication networks.The large-scale deployment of multiple wireless network access technologies and the multi-interface configuration of mobile terminal devices can effectively alleviate this conflicting demand.To give full play to the potential advantages of multi-interface communication devices,Multipath TCP(MPTCP)has emerged to support multiple parallel data transmission and is highly coupled with existing mainstream transmission protocols,which can effectively improve the transmission performance of mobile communication networks and provide users with better and more reliable network services.However,with the future development trend of network ubiquity and the evolution of network attack techniques,the network security situation is becoming more complex and severe.When a path in the Multipath TCP transmission system fails due to unexpected factors such as random link failures or deliberate network attacks,the degradation of the transmission performance of the path can negatively affect the robustness of the entire transmission system,which in turn affects the normal communication of the Multipath TCP transmission system.Therefore,the issues of Multipath TCP security and robustness deserve attention and discussion from researchers,but there is a serious lack of research on the issues related to Multipath TCP robustness and anti-destructiveness,and no uniform robustness measure has been identified.To address the above issues,the research in this paper consists of the following two main parts:(1)To enrich the existing research on Multipath TCP anomaly detection,this paper proposes a Multipath TCP anomaly traffic detection method based on signal analysis techniques.The network data flow in Multipath TCP transmission system is treated as a signal,and the signal characteristics of network attack anomalous traffic are mined at multiple scales by using wavelet analysis method in signal analysis technique,so as to identify and detect the anomalous traffic.Meanwhile,the accuracy metrics such as wavelet energy entropy and reconstruction error are calculated based on the wavelet decomposition and reconstruction results to measure the effectiveness of the anomalous traffic detection method.(2)To efficiently measure Multipath TCP transmission robustness,this paper proposes a cascade failure-based Multipath TCP transmission robustness optimization scheme.Based on the load-capacity cascade failure model,the cascade failure behavior in Multipath TCP transmission system is visualized from the perspective of different attack strategies,and the efficiency function is used to further measure the robustness of Multipath TCP transmission system.At the same time,the Multipath TCP robustness analysis model is further optimized based on the optimal load redistribution method with node idle capacity ratio,and the performance of the Multipath TCP robustness model is evaluated in terms of average throughput,packet disorder and jitter rate,respectively.Finally,a Multipath TCP transmission scenario is simulated using the NS2 platform and simulation experiments are conducted.The simulation experiment results show that the Multipath TCP robustness model proposed in this paper can realize the identification and detection of abnormal traffic,effectively measure the robustness of multipath transmission system,and meet the high quality service demand of streaming media applications while improving the reliability and robustness of multipath transmission system,which provides a strong driving force and unlimited potential for the continuous innovation of multipath transmission robustness research in heterogeneous networks.